ignore/remove NA values in read.csv

Question

I have a csv file as shown below that I read into R using read.csv, where column C has 12/30 empty values. I want to work out the max of each column, but the R function "max" returns "NA" when used on column C. How do I get R to ignore the empty/NA values, I cannot see an "rm.na" in read.csv?

data<-data.frame(read.csv("test.csv"))

data

A   B   C   
1   5   6
15  2   3
8   3   3
7   5   4
5   3   8
4   1   4
5   3   4
2   2   10
4   3   8
6   5   2
1   4   4
10  8   4
0   6   0
7   3   8
5   3   3
13  12  13
6   0   0
0   0   2
5   2   NA
7   3   NA
1   8   NA
11  1   NA
1   4   NA
0   7   NA
4   5   NA
3   10  NA
2   0   NA
6   4   NA
0   19  NA
1   5   NA

> max(C)
[1] NA

why not to use `max(..., na.rm = TRUE)`? – agstudy Apr 04 '13 at 10:15 — agstudy, Apr 04 '13 at 10:15

score 16 · Answer 1 · answered Nov 12 '13 at 09:06

16

    data<-na.omit(data)

then

    max(data)

If you do not wish to change the data frame then

    max(na.omit(data))

answered Nov 12 '13 at 09:06

Anurag Priyadarshi

1,113
9
17

score 9 · Accepted Answer · answered Apr 04 '13 at 10:48

9

you have two options that i can think of

 apply(data,2,max,na.rm=TRUE); # this will remove the NA's from columns that contain them

OR

apply(na.omit(data),2,max); ## this will remove the NA rows from the data frame and then calculate the max values

answered Apr 04 '13 at 10:48

Aditya Sihag

5,057
4
32
43

score 1 · Answer 3 · answered Apr 04 '13 at 10:50

1

I'd suggest to remove the NA after reading like others have suggested. If, however, you insist on reading only the non-NA lines you can use the bash tool linux to remove them and create a new file:

grep -Ev file_with_NA.csv NA > file_without_NA.csv

If you run linux or mac, you already have this tool. On windows, you have to install MinGW or Cygwin to get the tools.

answered Apr 04 '13 at 10:50

Paul Hiemstra

59,984
12
142
149

For me NA must be placed before filename: "grep -Ev NA file_with_NA.csv > file_without_NA.csv" – Doc Brown Dec 14 '16 at 14:34

score 1 · Answer 4 · edited Dec 21 '15 at 19:58

1

You should be able to use

max(x,na.rm=TRUE)

edited Dec 21 '15 at 19:58

NathanOliver

171,901
28
288
402

answered Dec 21 '15 at 14:23

Jess2332

11
1

ignore/remove NA values in read.csv

4 Answers4